Captions: Accessible Video for Deaf Viewers
Also called: Closed captions, Subtitles for the deaf and hard of hearing, SDH
Definition
Captions are synchronized text versions of a video's audio: the dialogue, who is speaking and the non-speech sounds needed to follow it. WCAG 1.2.2 requires them at Level A for all prerecorded audio in synchronized media.
Captions put the sound of a video on screen as text, in sync with the picture: what people say, who says it and the sounds that matter, such as "[doorbell rings]" or "[applause]". For deaf and hard-of-hearing viewers they are the only way to follow a product video, a webinar recording or a tutorial. They also help everyone who watches on mute – on the train, in an open office or late at night.
What it means
WCAG 1.2.2 Captions (Prerecorded) (Level A, WCAG 2.2) says: "Captions are provided for all prerecorded audio content in synchronized media, except when the media is a media alternative for text and is clearly labeled as such." Synchronized media means audio and video together – a typical video with sound.
According to W3C, captions convey what the audio track provides, so they include more than the spoken words:
| Part of the audio | Belongs in the captions? |
|---|---|
| Dialogue and narration | Yes, every line |
| Who is speaking, when it isn't obvious | Yes, e.g. "MARIA: …" |
| Sounds needed to understand the content | Yes, e.g. "[phone rings]", "[laughter]" |
| Background music without meaning | Optional, a short note is enough |
The exception is narrow: a video that only repeats text already on the page, and is clearly labeled as such, needs no captions. W3C failure F75 covers the opposite – a video without captions that shows more than the page says.
Live video is a separate criterion (1.2.4 Captions (Live), Level AA). People who cannot see the video need an audio description instead, and audio-only content such as a podcast needs a transcript.
Examples
<video controls>
<source src="/media/product-tour.mp4" type="video/mp4">
<track kind="captions" src="/media/product-tour.en.vtt" srclang="en" label="English" default>
</video>WEBVTT
00:00:01.000 --> 00:00:04.000
MARIA: Welcome to the new dashboard.
00:00:04.500 --> 00:00:06.000
[notification sound]The track element with kind="captions" (technique H95) adds closed captions to an HTML5 video; MDN documents the WebVTT format. Embedded players such as YouTube or Vimeo offer their own caption upload – the WCAG rule applies to them just the same.
Caption every prerecorded video with sound, including social clips and webinar recordings. Name speakers, describe meaningful sounds, keep timing in sync and check automatic captions by hand before you publish.
Publish raw auto-generated captions with wrong names and missing punctuation, leave out sound effects that carry meaning, or rely on a text alternative under the video instead of captions inside it.
How to check captions
Questions
Are auto-generated captions enough for WCAG?
Only if they are correct. Automatic captions often miss words, names, punctuation and speaker changes, and they rarely describe sounds. W3C failure F8 is captions that omit dialogue or important sound effects, so edit them before publishing.
What is the difference between captions and subtitles?
Subtitles usually render the dialogue, often as a translation, for viewers who can hear. Captions also identify speakers and describe important sounds, because they are written for people who cannot hear the audio.
Do videos without speech need captions?
1.2.2 covers audio content in synchronized media. If the sound carries information, such as an alarm or a spoken instruction, caption it. A video with only decorative music needs at most a short note.
Do I need captions under the European Accessibility Act?
The European Accessibility Act relies on EN 301 549, which takes over WCAG for websites. 1.2.2 is Level A there (clause 9.1.2.2), so prerecorded videos with sound on websites in its scope need captions.
Related terms
Transcript: Text Alternative for Audio and Video
A transcript is a text version of time-based media: the spoken words, who says them, important sounds and, for video, what happens on screen, in the right order. WCAG 1.2.1 requires such an alternative at Level A for prerecorded audio-only and video-only content.
Read definition: Transcript: Text Alternative for Audio and VideoAudio Description: Making Video Visuals Audible
Audio description is a spoken narration of a video's important visual content – actions, people, scene changes and on-screen text – placed in the pauses of the dialogue. WCAG 1.2.5 requires it at Level AA for all prerecorded video in synchronized media.
Read definition: Audio Description: Making Video Visuals AudibleAlt text
Alt text is a short written description of an image. Screen readers read it aloud, and browsers show it when the image can't load – so people who can't see the image still get its meaning.
Read definition: Alt textDeafness: How Deaf People Use the Web
Deafness is a substantial, uncorrectable hearing loss in both ears. Deaf people use websites without relying on sound, so speech and other audio must also be available as text – captions, transcripts – and, for some users, in sign language.
Read definition: Deafness: How Deaf People Use the Web